Breaking

Anthropic has closed a $65 billion funding round, pushing its valuation to $965 billion. That makes the maker of Claude the most valuable AI company on the planet, overtaking its better-known rival. TechCrunch reports this is likely the company's final private fundraise before an IPO.

The numbers behind the raise are striking. Annualized revenue has hit $47 billion, driven largely by enterprise adoption of its coding tools. Anthropic says the capital will go into safety research, computing capacity, and expanding its Claude product line. For any business currently evaluating AI vendors, Anthropic is no longer the underdog.

THE PLAY: Test Claude's Deep Research on a live business question this week, note where it delivers and where it fails, and fold that directly into your AI adoption thinking.

Follow the money out of the door

An unnamed company reportedly burned through half a billion dollars in Claude licensing fees in one month, according to The Decoder. The cause: no usage limits were set. No governance. No controls. Just an open tap.

This is not an edge case. It is a preview of what happens when procurement decisions run ahead of AI governance frameworks. If your organisation is deploying AI at scale without usage caps, cost monitoring, and clear accountability for spend, this story is a warning. The technology does not self-regulate. Your finance team will not thank you when the invoice arrives.

THE PLAY:Before you scale any AI deployment, set usage caps, assign budget ownership, and build cost monitoring in from day one. Governance is not a blocker to AI adoption. The absence of it is.

AI workforce adoption

Amazon built an internal ranking system to encourage AI adoption across the business. Employees responded by running meaningless AI tasks to inflate their scores, driving up the company's own cloud costs in the process. The Financial Times broke the story. Amazon has now killed the leaderboard entirely.

THE PLAY:he lesson for any leader rolling out an AI adoption programme is direct: measure outcomes, not activity. If you reward usage, you get usage. If you reward results, you get results. The two are not the same.

Anthropic confirmed a landmark funding round led by Google, with participation from Spark Capital and existing backers. The capital will primarily fund compute infrastructure and safety research.

Why it matters

The frontier AI race is now a capital-intensive infrastructure play. Anthropic, Google DeepMind, and OpenAI are collectively burning billions on compute. For enterprise buyers, this means more stable roadmaps — and higher switching costs.

Gartner's revised forecast puts global AI infrastructure investment — data centres, chips, networking — at $320 billion this year, up from $218 billion in 2025. The lion's share flows to hyperscalers, but enterprise spend on private AI infrastructure is growing fastest at 67% YoY.

African AI startups raised $480 million in Q1 2026, up from $160 million in Q1 2025, according to Partech Africa. Lagos, Nairobi, and Cairo are emerging as the primary hubs, with fintech-AI hybrids and agri-AI attracting the most capital.

Nigeria is punching above its weight. If you're building in Nigeria or looking to invest, the window for early positioning is right now — valuations remain a fraction of equivalent Western-market companies.

What it does: Anthropic has released Claude Opus 4.8, which now catches its own coding errors four times more often than its predecessor, according to The Decoder. More significantly for business users, it ships with a new Dynamic Workflows feature that can spin up hundreds of parallel sub-agents simultaneously, enabling tasks such as large-scale document reviews or codebase migrations to run automatically with limited human oversight.

THE PLAY: If you run any process that currently requires a team to work through large volumes of documents, data, or code in sequence, Opus 4.8's parallel agent capability is worth a direct test. Pick one bottleneck, run it through Dynamic Workflows, and measure time saved against your current baseline.

What it does: Runs quietly in the background on any video call, producing structured summaries with action items — no bot joins the call.
Best for: Exec teams tired of clunky Otter integrations.
Cost: Free tier available; Pro at $18/month.

THE PLAY:Turn on Granola for your next board or investor meeting. After, compare its action item list to your manually-taken notes. The gaps will tell you exactly what you're missing in real time.

Two stories this week, taken together, should concern every executive responsible for an AI budget. At Amazon, an internal programme designed to boost AI adoption was quietly undermined by employees gaming the metrics. The leaderboard measured usage rather than value, so staff ran pointless tasks to climb the rankings, costing the company money and producing nothing. Amazon has now scrapped the system entirely.

Separately, an unnamed enterprise reportedly spent $500 million in a single month on Claude licences after failing to set spending controls. At the same time, ClickUp cut 22% of its entire workforce to fund AI agent deployment, a decision Box founder Aaron Levie described this week as "AI psychosis" — the tendency of senior leaders to believe AI can replace functions they do not fully understand.

The pattern across all three stories is the same: AI spend running ahead of AI governance. Most organisations now have an AI strategy. Far fewer have the internal controls, measurement frameworks, and accountability structures to ensure that strategy produces a return. The gap between ambition and execution is where the real cost of AI lives. The question for your leadership team is not whether you are using AI. It is whether anyone is actually watching what it costs and what it delivers.

Read the full briefing at centaistrategy.com

Before you ask AI to produce anything complex, try starting with this instruction: "Before you begin, ask me all the questions you need to do this as well as possible." Then stop. Let the AI interrogate you.

Most people write a prompt and hope for the best. The result is a response built on assumptions. When you flip the dynamic and let the AI identify the gaps in your brief before it starts, the output quality improves significantly. You get a response shaped around what you actually need rather than what the model guessed you needed.

This works particularly well for strategy documents, communications, and any task where context and nuance matter. It takes thirty seconds longer at the start and saves considerably more at the editing stage.

Thanks for reading, more tomorrow, same time — The wAIve Daily Team

Found this useful? Forward it to someone who needs it.

Tell us what you need to know. Reply with your questions.